Prediction Based Policies for Efficient Selection of Cloud Data Center

Cloud computing has a variety of computing resources to facilitate the execution of large-scale tasks. Therefore, to select appropriate data center for executing a task is an important parameter to improve the performance of cloud computing environment. In present condition, available policies for data center selection are Round-robin, least connection, weighted least connection and so on. But they have some disadvantages like they did not focus on data centers load as well as weight has fixed value depending on the performance of data center. So we require more advance policies to improve the performance. In cloud environment, the response time observed at client side is mainly depends on load balancing algorithms and cloud data center selection technique. In this work we are proposing the data center selection prediction based techniques which predict the response time based on the history of response time for respective data center in the cloud and the selection of data center for upcoming request is done based on minimum predicted response time for each corresponding data center. Prediction based data center selection policy leads to effective reduction in response time observed at client side.